Keywords associated with document categories

Abstract

A system extracts a pair that includes a keyword candidate and information associated with a document from multiple documents, and calculates a frequency that the keyword candidate appears in search queries and a frequency that the pair appears in the multiple documents. The system also determines whether the keyword candidate is a keyword for a category based on the calculated frequencies, and associates the keyword with the document if the keyword candidate is the keyword for the category.

What is claimed is: 
     
       1. A method comprising:
 identifying, by one or more processors and as a keyword candidate, a first word included in one or more of previously submitted queries; 
 creating, by the one or more processor, a pair that includes the first word and a second word that differs from the first word,
 the second word being descriptive of a particular category of a plurality of categories; 
 
 calculating, by the one or more processors, a first value based on a first frequency that the first word appears in the previously submitted queries; 
 calculating, by the one or more processors, a second value based on occurrences of the pair in a collection of documents,
 one or more documents, in the collection of documents, being classified in the particular category based on occurrences of the second word in the one or more documents, and 
 the second value being calculated based on a second frequency that the first word appears in the one or more documents classified in the particular category; 
 
 determining, by the one or more processors and based on the first value, that the first value satisfies a first threshold; 
 determining, by the one or more processors and based on the second value, that the second value satisfies a second threshold; 
 determining, by the one or more processors, that the first word is a keyword for the particular category based on the first value satisfying the first threshold and the second value satisfying the second threshold; and 
 associating, by the one or more processors, the first word with the one or more documents based on determining that the first word is the keyword for the particular category. 
 
     
     
       2. The method of  claim 1 , where associating the first word with the one or more documents includes:
 indexing the one or more documents to include a reference to the first word. 
 
     
     
       3. The method of  claim 1 , where identifying the first word includes:
 identifying the first word when at least a particular number of user selections of search results, responsive to the one or more of the previously submitted queries, relate to the particular category. 
 
     
     
       4. The method of  claim 1 , where determining that the first word is the keyword for the particular category further includes:
 calculating a third value based on a third frequency that the first word appears in the collection of documents; and 
 determining that the first word is the keyword for the particular category further based on determining that the third value does not satisfy a third threshold. 
 
     
     
       5. The method of  claim 1 , where determining that the first word is the keyword for the particular category further includes:
 determining a score for the first word; and 
 determining that the first word is the keyword for the particular category further based on determining that the score satisfies a score threshold. 
 
     
     
       6. The method of  claim 5 , where the score for the first word is based on a third value associated with a third frequency that the first word appears in the collection of documents. 
     
     
       7. The method of  claim 1 , further comprising:
 receiving a search query that includes a form of the first word; and 
 increasing, in an ordered list of documents associated with the search query, a ranking of a document associated with the particular category. 
 
     
     
       8. The method of  claim 7 , where the ordered list of documents includes search results, and where the method further includes:
 outputting the search results after increasing the ranking of the document. 
 
     
     
       9. A non-transitory computer-readable medium storing instructions that are executable by one or more processors, the instructions comprising:
 one or more instructions to identify, as a keyword candidate, a first word included in one or more of previously submitted queries; 
 one or more instructions to create a pair that includes the first word and a second word that differs from the first word,
 the second word being descriptive of a particular category of a plurality of categories; 
 
 one or more instructions to calculate a first value based on a first frequency that the first word appears in the previously submitted queries; 
 one or more instructions to calculate a second value based on occurrences of the pair in a collection of documents,
 one or more documents, in the collection of documents, being classified in the particular category based on occurrences of the second word in the one or more documents, and 
 the second value being calculated based on a second frequency that the first word appears in the one or more documents classified in the particular category; 
 
 one or more instructions to determine, based on the first value, that the first value satisfies a first threshold; 
 one or more instructions to determine, based on the second value, that the second value satisfies a second threshold; 
 one or more instructions to determine that the first word is a keyword for the particular category based on the first value satisfying the first threshold and the second value satisfying the second threshold; and 
 one or more instructions to associate the first word with the one or more documents based on determining that the first word is the keyword for the particular category. 
 
     
     
       10. The non-transitory computer-readable medium of  claim 9 , where the one or more instructions to associate the first word with the one or more documents include:
 one or more instructions to index the one or more documents to include a reference to the first word. 
 
     
     
       11. The non-transitory computer-readable medium of  claim 9 , where the one or more instructions to identify the first word include:
 one or more instructions to identify the first word when at least a particular number of user selections of search results, responsive to the one or more of the previously submitted queries, relate to the particular category. 
 
     
     
       12. The non-transitory computer-readable medium of  claim 9 ,
 where the instructions further comprise:
 one or more instructions to calculate a third value based on a third frequency that the first word appears in the collection of documents, and 
 
 where the one or more instructions to determine that the first word is the keyword for the particular category include:
 one or more instructions to determine that the first word is the keyword for the particular category further based on the third value not satisfying a third threshold. 
 
 
     
     
       13. The non-transitory computer-readable medium of  claim 9 , where the one or more instructions to determine that the first word is the keyword for the particular category include:
 one or more instructions to determine a score for the first word, and 
 one or more instructions to determine that the first word is a keyword for the particular category further based on the score satisfying a score threshold. 
 
     
     
       14. The non-transitory computer-readable medium of  claim 13 , where the score for the first word is based on a third value associated with a third frequency that first word appears in the collection of documents. 
     
     
       15. The non-transitory computer-readable medium of  claim 9 ,
 where the instructions further comprise:
 one or more instructions to receive a search query that includes a form of the first word; and 
 one or more instructions to increase, in an ordered list of documents associated with the search query, a ranking of a document associated with the particular category, and 
 
 where the ordered list of documents includes search results. 
 
     
     
       16. A system comprising:
 a memory to store a plurality of instructions; and 
 a processor to execute the plurality of instructions in the memory to:
 identify, as a keyword candidate, a first word included in one or more of previously submitted queries, 
 calculate a first value based on a first frequency that the first word appears in the previously submitted queries, 
 create a pair that includes the first word and a second word that differs from the first word,
 the second word being descriptive of a particular category of a plurality of categories, 
 
 calculate a second value based on occurrences of the pair in a collection of documents,
 one or more documents, in the collection of documents, being classified in the particular category based on occurrences of the second word in the one or more documents, and 
 the second value being calculated based on a second frequency that the first word appears in the one or more documents classified in the particular category, 
 
 determine, based on the first value, that the first value satisfies a first threshold; 
 determine, based on the second value, that the second value satisfies a second threshold, 
 determine, based on the first value satisfying the first threshold and the second value satisfying the second threshold, that the first word is a keyword for the particular category, and 
 store information associating the first word with the one or more documents, associated with the particular category, based on determining that the first word is the keyword for the particular category. 
 
 
     
     
       17. The system of  claim 16 , where the processor, when associating the first word with the document, is further to:
 index the one or more documents to include a reference to the first word. 
 
     
     
       18. The system of  claim 16 , where the processor, when identifying the first word, is further to:
 identify the first word when at least a particular number of user selections of search results, responsive to the one or more of the previously submitted queries, relate to the particular category. 
 
     
     
       19. The system of  claim 16  where the processor is further to:
 calculate a third value based on a third frequency that the first word appears in the collection of documents, and 
 where the processor, when determining that the first word is the keyword for the particular category, is further to:
 determine that the first word is the keyword for the particular category further based on the third value not satisfying a third threshold. 
 
 
     
     
       20. The system of  claim 16 , where the processor, when determining that the first word is a keyword for the particular category, is further to:
 determine a score for the first word, and 
 determine that the first word is the keyword for the particular category further based on the score satisfying a score threshold. 
 
     
     
       21. The system of  claim 20 , where the score for the first word is based on a third value associated with a third frequency that the first word appears in the collection of documents. 
     
     
       22. The system of  claim 16 ,
 where the processor is further to:
 receive a search query that includes a form of the first word, and 
 increase, in an ordered list of documents associated with the search query, a ranking of a document associated with the particular category, and where the ordered list of documents includes search results. 
 
 
     
     
       23. The system of  claim 22 , where the processor, when increasing the ranking of the document, is further to:
 organize the search results, associated with the search query, based on increasing the ranking of the document associated with the particular category.